REMARKS ON A CERTAIN HYPOTHESIS IN SINGULAR PERTURBATIONS